analysis of multi temporal satellite imagery for total
TRANSCRIPT
International Journal of Environment and Geoinformatics 3 (1), 1-11 (2016)
Analysis of Multi Temporal Satellite Imagery for Total Suspended
Sediments in a Wave-Active Coastal Area-Gaza Strip Coastal Water,
Palestine
Midyan D.I. Aldabash1, Füsun Balık Şanlı1,*
1 Yıldız Technical University, Geomatic Engineering Department, 34220-Esenler İstanbul Turkey
Corresponding author*E-mail: [email protected]
Received 20 January 2016 Accepted 29 February 2016
Abstract
Sediment load materials is one of the key factors that determine the surface water quality, both of oceanic and
river water, and it specifies water optical properties. Thus it provides a background for a plenty of applications
and projects in the water and oceanography community. Landsat detects and classifies reflected solar energy
from bodies on the earth's surface. Suspended sediments existing in water column have an optical influences. So
that, Landsat images could detect suspended sediments concentration in such a water surface. In this study we
have three main objectives to be achieved as; TSS Concentration maps generation in the Gaza Strip coastal zone,
achieving analysis processes on TSS trend itself and TSS related coastal phenomenon, and investigation of the
ability of Landsat images to detect TSS comprehensively in a wavy coastal zone. For this purpose two landsat
TM5 images acquired in 1999 and 2010, one Landsat TM7 images acquired in 2003, and 2 Landsat Oli 8 images
acquired in 2014 and 2015 were used for TSS mapping. In addition, 64 TSS in-situ tested samples were also to
calculate a correlation equation between Digital Numbers - DN in each image pixels and TSS values in the
ground data. All image analysis and remote sensing steps have been done in this study using Integrated Land and
Water Information System - ILWIS software version ILWIS academic 3.3. Green and Red bands in all used
Landsat images contained the highest linear correlation factors -R- for the images acquired in 1999, 2003, 2010,
2014, and 2015. Resulted correlation factors were higher by reducing time difference between acquisition time
and sampling time. Generated maps showed that circulation in Gaza coastal area are counterclockwise, and it
brings the sediments from Nile River Delta toward Gaza Strip.
Keywords: TSS, Landsat, circulation, water column
Introduction
Gaza Strip's economy is highly dependent upon
recreational and touristic nature, and fishing
activities in its coastal zone in Mediterranean.
Based upon 2003 report of PCHR (Palestinian
Centre for Human Rights) and the International
Commission of Jurists - Geneva, fishing industry
provides 2500 job opportunity and it operates
727 fishing boats, and produces 2309 metric ton
of fish.
Gaza Strips' beach has a wealth recreational
value for Gazians. Isolated from world as both
Israel and Egypt tightly control Gaza's border,
many residents turn to the beach as a temporary
escape. It also provides an opportunity for
residents of the territory to set up businesses with
a relatively small financial investment (United
Nations - UN, 2014).
Suspended solids reduce visibility and absorb
light, which can increase stream temperatures
and reduce photosynthesis. Impeding aquatic
plant photosynthesis reduces the amount of food,
habitat and dissolved oxygen available for other
species. Fine particles may also clog and abrade
fish and insect gills and tissue and, interfere with
egg and larval development. Pollutants such as
pesticides and PCBs adhere to the surfaces of
TSS and can be transported into aquatic
environments in this fashion [1].
Healthy coastal ecosystem is very crucial for
Gaza's economy and coastal activities. Coastal
ecosystem parameters must be monitored
1
M.D.I. Aldabash and F. Balık Şanlı / IJEGEO 3 (1), 1-11 (2016)
2
repetitively and comprehensively. Accurate
estimate of water quality concentrations is also
vital for determining the health of ecosystem
(Guang 2009, Karr and Dudley 1981). Total
Suspended Sediments - TSS is one of these
parameters that must be always monitored in
coastal surface.
Most sediments in surface water derives from
surface erosion and comprises a mineral
component, arising from erosion of bedrock and
an organic component arising from soil forming
processing ( E. Ongley, UNEP\WHO, 1996).
While historical research on turbidity and
suspended sediments has been thorough, studies
of stream bed sedimentation have typically
relied on semi-quantitative measures such as
embeddedness or marginal pool depth (Mark S.
Riedel, 2006). In Palestine, data about
environmental and aquatic parameters like TSS
are normally based on an individual shipboard
from different points in the aquatic surface, in
other words, it is non-comprehensive and does
not provide a complete vision for TSS transport
and deposition.
These are followed by laboratory measurements
of the water constituents and interpolations
between sampling stations to create maps
illustrating the spatial distribution of the
parameters (Pennock, 1985). This process is
time and cost consuming and involves a
reasonably high degree of inaccuracy
.Moreover, all of the previous studies in TSS
mapping were in an wavy non-active areas like
rivers, valleys, deltas, or far shore oceanic areas,
but this study is the first in a wavy active coastal
zone, at least in Gaza Strip.
Since remote sensing has enormous spatial and
temporal operative abilities and techniques, it
could be an ideal alternative to the shipboard in-
situ, and can overcome a lot of its problems.
Water quality mapping by using remote sensing
provides better results at a relatively cheaper
cost (Wheather Bee and Kleams, 1998).
In this study, 64 tested TSS samples from
different locations along mid shore, near shore,
and far shore of Gaza coastal zone were
correlated with their corresponding digital
number in each acquired Landsat images for five
years. After that, correlation equations have
been applied on each images using ILWIS
remote sensing software to produce
concentration maps for each image.
Brief Review of Landsat Applications in TSS
Mapping.
Landsat satellites have repetitive, circular, sun-
synchronous, near polar orbits, providing full
coverage between 81° N and 81° S. The sensors
always scan the ground at satellite nadir.
(Eurimage, Landsat, 2013). The remote sensor
signals results from the interaction between solar
radiation and both water and atmosphere (Kirk,
1986).
Pure water presents high absorption in the red
and infrared region in the electromagnetic
spectrum. Total suspended solids (TSS) describe
particulates of varied origin, including soils,
metals, organic materials and debris that are
suspended in a moving body of water.
Turbulence keeps the particulates suspended in
water allowing the solids to be transported
downstream. The water spectral response is
changed by their optically active components
such as pigments, organic and inorganic matter,
and organic dissolved substances (Maycira
Pereira de Faras Costa, 1998).
Suspended inorganic matter are the main light
scatters within the aquatic environment. Size,
shape, and concentration are the main factors
explaining the amount of scattering by inorganic
matter (Novo et al., 1989).
Using this relation between substances
concentration in water column and the scattered
and reflected energy toward the satellite's sensor,
statistics and analysis techniques in remote
sensing find the way for TSS concentration
detection in Gaza coastal zone.
Study Area
Gaza Strip (Qita'a Gazza), is a narrow area
located on the Southern part of Palestinian
coastal plains, and surrounded by Sinai
Peninsula from South, Negev Desert from East,
Israel from North, and Mediterranean Sea from
the west. Gaza Strip is located between
longitude 34° 20" to 34° 25" East and 31° 16"
and 31° 45"North (Fgure 1). Gaza has 360 km2
M.D.I. Aldabash and F. Balık Şanlı / IJEGEO 3 (1), 1-11 (2016)
3
gross area with 72 km2 as a border area, 59 km2
of this area has been made by Israel while the
other 13 km2 has been made by Egypt. Gaza
Strip is 41 km long on the coast, 7 km wide from
North, and 12 km wide from South along Sinai
Peninsula.
According to the statistical surveying that was
done in June 2015, the population of Gaza is
more than 1,85 million inhabitant [1].
Population density in the Gaza Strip renowned
as among the highest in the world-rose about
38.5 % in 1997, to an average of 3880 people per
km2 . Gaza Strip has five governorates. The
central one is Gaza City, while the largest one is
Khanyunus in the South. Rest of the three
governorates are North of Gaza, Deir Albalah,
and Rafah (UNRWA, 2007).
Fig.1. Location of Gaza Strip
Average daily temperature in Gaza Strip ranges
from 26 C° in Summer to 12 C° in Winter with
the average daily maximum temperature ranges
from 29 C° to 17 C°, and the minimum
temperature ranges from 21 C° to 9 C° in
Summer and Winter respectively (GMS, 2009).
Rainfall occurs in the winter period between
October and March, while the winter period
from June to September is dry with no rainfall.
The average rainfall varies between 200-400
mm. (PMO, 2008)
Materials and Methods
In this study two Landsat TM5 imagery, one
Landsat TM7 imagery, and two Landsat Oli
imagery were used. Landsat is an American
satellite that has been launched by NASA, and
recently they launched the last version of the
satellite, Operational Land Imager - Lansat 8
Oli. Image data are nine spectral bands (Band
Designation) with a spatial resolution of 30
meters for Band1 through Band7, while Band 8
(Panchromatic) is 15 meters. Scene size in Oli is
170 Km N-S by 183 km E-W (SEW, 2015).
Data format in Oli is GeoTiff extension with 16-
bit pixel value, while the map projection is
Universal Transverse Mercator - UTM (Polar
Stereographic for Antarctica) and the datum is
WGS 84. We used 64 TSS tested and
coordinated samples that were collected from
three different shorelines, near, mid, and far
shores from the Gaza coastal zone. These data
were granted by the department of Applied Civil
Engineering in University of Palestine - Gaza,
and tested using the "Crucible Dishes" method
using "Schleicher & Schuell" with pore size 62
micron and 110 mm in diameter.
ILWIS Input Data
ILWIS is an acronym for the Integrated Land
and Water Information System. It is a
geographic information systems - GIS with
image processing capabilities. ILWIS has been
developed by the international institute for
Aerospace Survey and Earth Sciences - ITC,
Enschede-Netherlands.
From the date you can generate information on
the spatial and temporal patterns and processes
on the earth surface (ILWIS ITC user guide
book).
M.D.I. Aldabash and F. Balık Şanlı / IJEGEO 3 (1), 1-11 (2016)
4
Used imageries in this study were taken in
different dates as shown Table1. Furthermore,
all ground data were collected through two
marine tours in 2012 winter as shown in Table 2.
Samples sites are all coordinated using manual
GPS on the UTM projection and WGS 84 datum.
Delivered images are all free of clouds and haze
and they were georeferenced same as the
samples (UTM and WGS 84).
Table 1. Used Landsat images and corresponding acquisition date
Landsat Version Acquisition Date
Landsat TM5 11.02.1999
03.12.2010
Landsat TM7 20.10.2003
Landsat Oli 8 19.11.2014
22.01.2015
Fig 2. Illustrates Gaza Strip and its coastal zone for the five used landsat scenes in a) 1999, b) 2003,
c) 2010, d) 2014, and e) 2015.
Fig. 3. Illustrates ground data samples sites along Gaza coastal area.
Methodology
The methodology used in this study is consisting
six main steps beginning by data extraction and
culminating by slicing and layout.
Step1: Extracting each corresponding pixel
value (Digital Number - DN) for each sample
site in the Green, Red, Blue, and Near Infrared -
NIR bands in each imagery. Five tables were
generated and each table consisted four columns
(one for each band), and there was a 64 line or
cell in each column (one for each sample). This
step is completely done using ILWIS.
Step2: Statistical analysis was done for each
image to find a correlation value between each
band DN's values and the TSS values in tested
a b c d e
M.D.I. Aldabash and F. Balık Şanlı / IJEGEO 3 (1), 1-11 (2016)
5
samples. Highest correlation was being chosen
as the trusted correlation for each image.
In this process we found the linear regression
mathematical relation (Linear Equation) for the
highest correlation value as a function of DN as
input, and the output is TSS concentration values
in mg\l. Thus we had five equations as illustrated
in Table 3.
Step 3: Extraction Gaza zone from the whole
imagery of Palestine. This step was done by
specifying the number of columns and lines for
the Top-most-left and Bottom-most-right
corners pixels for each image. Each image
consists approximately 1450 pixels N-E and
1150 pixels W-E (total are more than 1.6 x 106
pixels).
Step 4: Land-Sea masking, is the process of
Gaza Sea water surface lineation to separate it
from land. NIR band was used to discriminate
between land and sea, because there is a high
difference in pixel values between the land and
sea features. Therefore a specific pixel value has
been selected as a threshold value for sea.
Writing a short code in ILWIS command
window can do the step. For instance, the NIR
band in 1999 image is named as (B4LSM99),
and the threshold pixel value is 30, here is the
code:
B4LMS99 = Iff(B4LMS99<=30, B4LMS99,
0)
Table 3. Correlation formulas and factors for each image
Acquisition
Year
Equation f(DN) Correlation Factor-R Band
1999 TSS = 1,790 DN - 14,27 0.068 Red
2003 TSS = 2,766 DN - 49,54 0.668 Green
2010 TSS = 2,855 DN - 45,51 0.700 Green
2014 TSS = 0,017 DN - 109,1 0.702 Green
2015 TSS = 0,048 DN - 288,6 0.644 Red
M.D.I. Aldabash and F. Balık Şanlı / IJEGEO 3 (1), 1-11 (2016)
6
Fig. 4. Illustrates correlation curves, equations, and factors for years a)1999, b)2003, c)2010, d)2014
and e)2015.
Fig. 5. Writing lineation code in ILWIS command window.
It means that keep any value less than or equal
30, and give zero for the rest of the values in the
layer band B4LMS99.
In the same step all the non-zero valued pixels
have been replaced by their corresponding pixels
in the bands that has the highest correlation value
(Green or Red). Generated bands are land-Sea
masked bands, and they are ready to continue the
required calculation on them.
Step5: Total Suspended Sediments calculations
were done by applying the regression equation
on the bands that had the highest correlation
values. For instance, in 2014 image, highest
correlation was on Green band, and the
regression equation was TSS = 0,017 DN -
109,1. Writing a code in ILWIS could perform
this calculation.
B314CSS = ((0.017xB314)-109.1)
M.D.I. Aldabash and F. Balık Şanlı / IJEGEO 3 (1), 1-11 (2016)
7
Fig. 6. a)Illustrates the used code in ILWIS command window for TSS concentrations calculations.
b) Resulted TSS map from the used equation (TSS=0.017DN-109.1) as an example.
Step 6: Classifying and slicing the resulted TSS concentration maps using Psedu colors graduation to
start analysis through maps.
a
b
a b c
M.D.I. Aldabash and F. Balık Şanlı / IJEGEO 3 (1), 1-11 (2016)
8
Fig. 7. Final layout for the generated TSS concentration maps from Landsat images scenes that were
acquired in a) 1999, b) 2003, c) 2010, d) 2010, and e) 2014.
Results and Discussion
The results are relatively near the expectation
and coincides for high degree with the other
studies.
1. Linear regression analysis was applied to data
from each sampling point. Results indicated that
the best correlation for 2003, 2010, and 2014
were on the Green bands (0.52 - 0.6 µm).
Highest correlation factor values in 1999 and
2015 images were at Red bands (0.64 - 0.67 µm).
Thus, the best bands for TSS mapping are Green
or Red bands. This variation in bands refers to
TSS composition, Organic and Inorganic
substances. If the dominant composition was
inorganic (Clay and sand), red band is expected
to contain highest correlation and vice versa.
Soil reflectance decreases as organic matter
increase [3].
2. Results showed that the highest TSS
concentration in all images are in Near-Shore
area (Near the beach), and concentrations is
inversely proportional with distance from beach.
That's because near-shore area is wavy active
and waves always suspense the Bed load in the
water column. Moreover, bathymetry depth in
near-shore area is low, so that reflectance is high.
Fig. 7. TSS concentration manner by going
toward far shores in the area.
3. Suspended sediment concentration increases
by going to the southern direction (Egypt)
toward Nile river Delta, which is considered the
main sediment discharging source in the
Mediterranean
d e
M.D.I. Aldabash and F. Balık Şanlı / IJEGEO 3 (1), 1-11 (2016)
9
Fig. 8. TSS concentration manner by going
toward south, to Egyptian side.
4. Circulation direction in Gaza coastal area is
counterclockwise as shown, especially in 1999
and 2003 maps. This visually explains the cause
of discharging sediments from Nile delta to Gaza
Area.
Fig. 9. Illustrates the counterclockwise direction
for currents circulation in the Gaza coastal area.
5. Since circulation in Gaza coastal area is
counterclockwise, Gaza seaport prevents
sediment passing toward the Northern of Gaza
seaport. This explains the cause of coastal
erosion in the area on the Gaza seaport northern
side, and the accretion on the southern side of
Gaza Seaport.
Fig. 10. a) TSS accumulation on one side and
lacking on the other side of Gaza seaport. b)
Beach erosion and accretion around Gaza
seaport as a result to TSS disturbance in
distribution around the seaport.
6. Straight correlation line slope in each
correlation model is positive. It means that the
higher TSS concentration (Less pure water) is
the higher energy reflectance, thus higher pixel's
DN.
7. Correlation factor values - R are inversely
proportional with the time difference between
time acquisition date and sampling date. Lower
time period between sampling data acquisition
date leads to higher correlation values.
8. Correlation factors-R for the 2010 and 2014
images were approximately the same, 0.700 and
0.702 respectively. This is because the time
duration between acquisition and sampling were
the same for both of the mages, 2 years.
9. Landsat images have been used effectively in
TSS mapping for serving coastal monitoring,
although an error margin is included. This errors
b
a
M.D.I. Aldabash and F. Balık Şanlı / IJEGEO 3 (1), 1-11 (2016)
10
could be through errors in DN as a result to
bottom reflection, sunlight, and\or atmospheric
conditions.
10. Suspended sediment concentrations depends
on 3 main factors. Firstly, TSS is high in rainy
season because of high flooding toward sea.
Secondly, TSS is lower in low tidal conditions
where higher water levels. Thirdly, waves
activity suspends bed load in water column, thus
increases the TSS concentrations. Therefore,
stronger and higher waves lead to higher TSS
concentration.
Conclusion
In this paper, Landsat images have been
effectively used in coastal monitoring regarding
mapping TSS concentration comprehensively in
Gaza coastal area as a wavy area. Where all
works that have been done in this sector were in
a non-wavy areas. Works included an error
margin for many reasons that I previously
mentioned. Simply, if those reasons are taken
into consideration, highest possible accuracy
could be achieved. Moreover, this is the first
study that shows TSS in Gaza coastal area
generally, and especially in a comprehensive
form under the siege and hard situations there.
All the coastal parameters that have been studied
agreed with the coastal studies that were done
previously, such as circulation and the eroded
areas around Gaza seaport. Furthermore, Gaza
seaport is big environmental problem and it must
be either modified by opening a holes in its walls
to allow sediment passing, or construct a new
one within Gaza land.
Higher TSS concentrations also affect the
aquatic marine life through reducing the
required light penetration amount for
photosynthetic processes for aquatic species.
Finally, TSS in Gaza coastal area is highly
influenced by surrounding environmental
factors and realities, such as rainfall, winds, tidal
conditions, etc.
Acknowledgments
Authors gratefully acknowledge to Prof. Dr.
Hassan Hamouda and Ass.Prof.Dr. Nadine Abu
Shaaban from University of Palestine for
providing the used ground data, and to Eng.
Abdallah Shahin from COMOSATS - Pakistan,
who helped too much in images downloading.
References
Bayer, J.M., and Schei, J.L., eds. (2009),
PNAMP Special Publication: Remote
Sensing Applications for Aquatic Resource
Monitoring. Pacific Northwest Aquatic
Monitoring Partnership, Cook, Washington,
100 p.
H. S. Lim, M. Z. MatJafri, K. Abdullah and
Robabeh Asadpour. (2011), TSS Mapping
Using THEOS Imagery over Penang Island,
Malaysia. School of Physics, Universiti
Sains Malaysia.
Hossam Adel Zaqoot, Taysir Saleem Hujair,
Abdul Khalique Ansari, Shaukat Hayat
Khan. (2010) , Assessment of Land-based
Pollution Sources in the Mediterranean Sea
along Gaza Coast. Mehran University of
Engineering and Technology, Jamshoro,
Sindh, Pakistan.
Javier A. Concha. (2014), The Use of Landsat 8
for Monitoring of Fresh and Coastal Water.
Digital Imaging and Remote Sensing Lab-
DIRS, Universidad de Concepción,
Uruguay.
Jean Parcher. (2012), Landsat History and
Legacy. UNCOPUOS Special Panel for the
40th Anniversary of Landsat, USGS.
Mahmoud Ali. (2002), The Coastal Zone of
Gaza strip-Palestine Management and
Problems. Mediterranean network to Assess
and upgrade the Monitoring and forecasting
Activity in the region (MAMA)-Paris.
Maycira Pereira de Far~as Costa. (1998),
Coastal Water Chlorophyll Estimation Using
Landsat TM. Instituto Nacional de Pesqulsas
Espaclals.
Mazen Abualtayef, Said Ghabayen, Ahmed Abu
Foul, Ahmed Seif, Masamitsu Kuroiwa,
Yuhei Matsubara, Omar Matar. (2012), The
Impact of Gaza Fishing Harbour on the
Mediterranean Coast of Gaza. IUG-Gaza,
CRI- Egypt, Tottori University, Japan.
Mazen Fadel El Bana, and Dr. Zeyad Abu Heen.
(2011), Sanitary and Social Effects of Dear
El Balah Landfill, Gaza Strip-Palestine. The
Islamic University –Gaza, Water Resources
Management.
Mohammad A. Rahman. , Bernd Rusteberg,
Mohammad S. Uddin, Muath A. Saada,
M.D.I. Aldabash and F. Balık Şanlı / IJEGEO 3 (1), 1-11 (2016)
11
Ayman Rabi, and Martin Sauter. (2014),
Impact Assessment and Multicriteria
Decision Analysis of Alternative Managed
Aquifer Recharge Strategies Based on
Treated Wastewater in Northern Gaza.
Leibniz University, Hannover, Germany,
and Georg-August Universität Göttingen,
Goldschmidtstr, Germany.
Mustafa Ustuner, Fusun Balik Sanli1, and
Barnali Dixon. (2014), Application of
Support Vector Machines for Landuse
Classification Using High-Resolution
RapidEye Images: A Sensitivity Analysis.
Department of Geomatic Engineering, Yildiz
Technical University, and Department of
Environmental Science, Policy and
Geography, University of South Florida.
Stephen E. Wood. (2015), Summary related to
course ESS 421 Introduction to Remote
Sensing. University of Washington.
U.S. Environmental Protection Agency, Office
of Solid Waste. (1999), Data Collection for
the Hazardous Waste Identification RULE,
Section 11.0: Aquatic Food Web Data. EPA,
Washington, DC.
Unit Geo Software Development Sector Remote
Sensing & GIS. (2001), ILWIS 3.0
Academic, User's Guide. International
Institute for Aerospace Survey and Earth
Sciences -ITC, Netherlands.
United Nations Environment Programme-
UNEP, and the World Health Organization-
WHO, Jamie Bartram and Richard Ballance,
E. Ongley. (1996), Water Quality
Monitoring - A Practical Guide to the Design
and Implementation of Freshwater Quality
Studies and Monitoring Programmes, Ch.13
Sediment Measurements.
United Nations for Relieving and Working
Agency-UNRWA, Salem Ajluni. (2010),
West Bank and Gaza Strip Population
Census of 2007. UNRWA Headquarter,
Jordan-Amman.
United States Geological Survey-USGS. (2012),
Landsat- A Global Land-Imaging Mission.
USGS EROS, Sioux Falls, USA.
World Bank, Daniela Gressani, A. David Craig.
(2007), West Bank and Gaza Investment
Climate Assessment: Unlocking the
Potential of the Private Sector. World Bank,
Social and Economic Development
Department ,Middle East and North Africa
Region.
Internet References
1. http://www.stormwaterx.com/Resources/Ind
ustrialPollutants/TSS.aspx
2. http://web.pdx.edu/~emch/ip1/bandcombina
tions.html
3. http://www.lakesuperiorstreams.org/underst
anding/param_turbidity.html
4. https://www.cia.gov/library/publications/the-
world-factbook/geos/gz.html